From Solvent-Free to Dilute Electrolytes: Essential Components for a

Dec 8, 2017 - ... Blaustein Institutes for Desert Research, Ben-Gurion University of the ...... and Dublin Philosophical Magazine and Journal of Scien...
0 downloads 0 Views 552KB Size
Subscriber access provided by READING UNIV

Letter

From solvent free to dilute electrolytes: Essential components for a continuum theory Nir Gavish, Doron Elad, and Arik Yochelis J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.7b03048 • Publication Date (Web): 08 Dec 2017 Downloaded from http://pubs.acs.org on December 11, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

From Solvent Free to Dilute Electrolytes: Essential Components for a Continuum Theory Nir Gavish,† Doron Elad,† and Arik Yochelis∗,‡ †Department of Mathematics, Technion - IIT, Haifa, 3200003, Israel ‡Department of Solar Energy and Environmental Physics, Swiss Institute for Dryland Environmental and Energy Research, Blaustein Institutes for Desert Research, BenGurion University of the Negev, Sede Boqer Campus, Midreshet Ben-Gurion 8499000, Israel ¶Department of Physics, Ben-Gurion University of the Negev, Be’er Sheva 8410501, Israel E-mail: [email protected]

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The increasing number of experimental observations on highly concentrated electrolytes and ionic liquids show qualitative features that are distinct from dilute or moderately concentrated electrolytes, such as self-assembly, multiple-time relaxation, and under-screening, which all impact the emergence of fluid/solid interfaces, and transport in these systems. Since these phenomena are not captured by existing mean field models of electrolytes, there is a paramount need for a continuum framework for highly concentrated electrolytes and ionic liquids. In this work, we present a self-consistent spatiotemporal framework for a ternary composition that comprises ions and solvent employing free energy that consists of short and long range interactions, together with a dissipation mechanism via Onsagers’ relations. We show that the model can describe multiple bulk and interfacial morphologies at steady-state. Thus, the dynamic processes in the emergence of distinct morphologies become equally as important as the interactions that are specified in the equilibrium-free energy. The model equations not only provide insights to transport mechanisms beyond the Stokes-EinsteinSmoluchowski relations but also enables to qualitative recovery in the full range (three distinct regions) of non-monotonic electrical screening length that has been recently observed in experiments using organic solvent to dilute ionic liquids.

Graphical TOC Entry

2

ACS Paragon Plus Environment

Page 2 of 22

Page 3 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Keywords electrolytes, ionic liquids, electrical screening, gradient flow, self-assembly, bifurcation theory, pattern formation

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Concentrated electrolytes are being examined for a broad range of applications, 1–4 and specifically for energy related devices, examples of which include dye sensitized solar cells, fuel cells, batteries and super-capacitors. 5–7 The optimized design of these devices requires the mechanistic spatiotemporal understanding of ionic arrangement and charge transport in electrolytes. Although the physicochemical aspects of electrolyte solutions have been extensively studied, a series of recent experimental and computational results 8–19 reflects knowledge gaps even in the context of basic science. 3,16,20–24 A robust spatiotemporal framework is thus required to advance electrochemically-based industrial applications of highlyconcentrated electrolytes, e.g., fuel cells ∼1M (mol/liter) and batteries ∼10M. Spatiotemporal theoretical formulation for electrolytes goes back to 1890’s where meanfield Poisson-Nernst-Planck (PNP) framework was originated. 25,26 The PNP model describes ions in electrolytes as isolated point charges which obey drift-diffusion transport under an electric potential, and indeed serves a basis for dilute electrolytes. 27 This description is valid for dilute electrolytes (below ∼ 0.01M), but is oversimplified for concentrated electrolytes where, for example, the finite size of the ions is important. 28,29 Consequently, extensive studies during the past centennial have led to major modifications of the mean-field approach to tackle the asymptotic electrical double layers (EDL) or electrokinetic properties, see representative reviews 20,21,24,30 and references therein. Among the broad family of modified PNP and Poisson-Boltzmann (PB) models, we highlight for our purposes the steric-PNP (SPNP) approach that has been employed for studying permeability and selectivity in ion channels, 31 where ion concentrations exceed 10M. This approach accounts for Lennard-Jones type interactions between the ions, which are similar to the interactions employed in molecular dynamic (MD) simulations. Theoretically (i.e., excluding the ion-pairing question 22 ) the upper limit of ‘concentrated electrolytes’ pertains to molten salts, which about room temperature are also being referred to as ionic liquids (IL), 5,6 i.e., solvent-free electrolytes. Besides the multiple applications, IL either with or without dilution also pose several physicochemical properties that

4

ACS Paragon Plus Environment

Page 4 of 22

Page 5 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

extend the phenomenology of concentrated electrolytes: (i ) multiple timescales, 17,32 (ii ) self-assembly, 33–35 and (iii ) non-monotonic variation (exhibiting qualitatively three distinct regions) of EDL with concentration, 15,36 a.k.a., the phenomenon of under-screening. These phenomenologies emphasize that besides the vast progress in concentrated electrolytes, a unified framework that combines altogether the above mentioned phenomena is still missing. 24 Here we present a dynamical and thermodynamically consistent, unified continuum framework for ternary media based on SPNP and Onsager’s relations. Our goal is to present a theoretical framework by outlining the essential features that are designed to pave the road toward a realistic description for concentrated electrolytes: (i ) explicit densities of electrically positive, negative and neutral subsets (aqueous or organic carriers), (ii ) finite-size (steric) effects, and (iii ) selective energy dissipation to reflect mass transport. While we follow the SPNP formulation, the starting point stems from IL 37,38 and introduce explicitly the solvent density. The model is qualitatively tested and appear to capture and relate, in broad parameter regimes, the above mentioned phenomena to effects of polarization 3 and self-assembly 23 to explain the origins of non-monotonic electrical screening length variation. 15,36 In particular, at high concentrations, the system is expected to indeed form different morphologies in both the EDL and bulk region that are strongly dictated by spatiotemporal instabilities that drive the self-assembly process in the bulk, as in models for pure ionic liquids 37,40 and charged polymer diblock copolymers. 41

Model equations for a ternary composition In general, PNP-type models consider the solvent as an effective background medium. This assumption, however, breaks in the high concentration limit, e.g., in ionic liquids. Respectively, we derive a ternary model by considering first model equations for asymmetric pure IL 38 and introducing the solvent explicitly. Specifically, we assume a monovalent electrolyte

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

comprising cations (p), anions (n) and solvent (s), identified by their the molar concentrations and that maintain uniform densities in terms of partial volume that is occupied by each species: p pmax

+

n nmax

+

s smax

= 1,

(1)

where pmax , nmax and smax being the maximal concentrations (packing densities) of the cations, anions and solvent molecules, respectively. Following Onsagers’ framework, cf. 37 , the transport equations are given by    p −p −p   p J   p  pmax − p     ∂   n  = ∇ ·  −n   n Jn  . n − n −n max     ∂t       s sJ s −s −s smax − s 



(2a)

and the Coulombic interactions obey the standard Poisson’s equation

∇ (∇φ) = q(n − p),

(2b)

where q is the elementary charge and  is the dielectric permittivity (assumed here to be constant). The respective fluxes are obtained from variation (functional derivative) of the free energy Jp =

µB δF ∇ , cmax δp

Jn =

µB δF ∇ , cmax δn

Js =

µB δF ∇ , cmax δs

where ˆ

entropy

electrostatic

z }| { z }| i{ h p n si h  2 F = dx kB T p ln + n ln + s ln + q(p − n)φ − |∇φ| c c s 2 Ω " !# 2 2 β κ2 p ∇ n + c¯max np + E0 ∇ + , nmax nmax pmax 2 pmax | {z } local approximation of Lennard-Jones interactions

6

ACS Paragon Plus Environment

(3)

Page 7 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Ω is the considered physical domain, µB is the mobility coefficient, β is the interaction parameter (a.k.a. Flory parameter in the context of mixing energy for polymers) for the anion/cation mixture, E0 κ2 is the gradient energy coefficient with units of energy for the former and units of length for the latter, and c¯ is the average concentration of ionic species and s¯ is the average concentration of solvent molecules, 1 c¯ := |Ω|

ˆ

1 p dx = |Ω| Ω

ˆ

1 s¯ := |Ω|

n dx, Ω

ˆ s dx. Ω

Note that global charge neutrality and the equal valence of anion and cations ensure that the average concentration of cations equals the average concentration of anion. This implies, together with the electrolyte incompressibility assumption (1), that the maximal average concentration (achieved when s¯ = 0), equals the harmonic average of pmax and nmax 1 c¯max

=

1 pmax

+

1 nmax

.

Model interpretation To uncover the physical properties the key parameters of the model equations (2), we rewrite the equations in their dimensionless form (see SI for details) and for simplicity present only the equation for cations while the equation for anions is given in the SI:  ∂p = ∇ · JpPNP + cJp1 + c2 Jp2 + c3 Jp3 , ∂t

7

ACS Paragon Plus Environment

(4)

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

where

JpPNP = ∇p + p∇φ, | {z }

(5a)

drift-diffusion

Jp1 =

2Υ χ pJpP N P − pJnP N P + p∇ (p + n) − p∇n, | {z } 1+γ 1+γ | {z } saturated mobility | {z } solvent entropy

Jp2 = −

(5b)

inter-diffusion

  1 p σ∇3 p + χ∇ (pn) , 1+γ {z } |

(5c)

steric effects

Jp3

 σ = p p∇3 p + n∇3 n . (1 + γ) | {z }

(5d)

steric effects

The normalized global ion concentration 0 ≤ c ≤ 1 is a key parameter that relates between the ions and the solvent via (1 − c) s = 1−c (p + n). In particular, c is scaled such that c = 0 corresponds to absence of ions (dilute limit), while c = 1 to absence of solvent (ionic liquid), i.e., powers of c describe equal contributions of terms in the equations and do not imply a perturbative expansion. Here, σ is a measure for the ratio between relative strength of short range steric and long range Coloumbic interactions, χ is also related to steric interactions rescaled to the ambient temperature, γ is the size ratio between cations and anions, and Υ is the complementary size asymmetry between the solvent and ions. In particular, Υ ∼ 1 for solvent molecules whose size is similar to the size of the ions (e.g., DMF), and Υ  1 for solvent molecules which are much smaller than the ions (e.g., water). It is instructive to consider the ionic flux term-by-term to reveal the impact of different mechanisms on electrolyte behaviour as concentration is varied. For simplicity of the exposition we do so by choosing Jp while components of Jn are given in the SI. At the dilute limit, c  1, steric effects are perturbative, i.e., Jp ≈ JpP N P . Therefore, equation (4) reduces to the classical PNP theory. Similarly, in the solvent-free limit, c ' 1, the model (4) reduces to the model for pure ionic liquids 38 that was shown to exhibit multiple timescale dynamics in the context of transient currents 37 . To understand how electrolyte behaviour varies at

8

ACS Paragon Plus Environment

Page 9 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

intermediate concentrations, let us consider initially, terms of Jp1 , that is, c contributions. The first term in Jp1 is related to solvent entropy that reflects the solvent tendency to diffuse to region of high ionic concentration (thus, low solvent concentration). This effect becomes dominant when Υ  1, i.e., when solvent molecules are significantly smaller than the ions, since entropy per volume of many small particles is larger than the entropy of fewer larger particles. The second term corresponds to saturated mobility, which gradually diminishes the transport of cations to cation-rich regions. This term prevents the formation of regions with a local concentration of constitutes beyond their packing density. The final two terms correspond to inter-diffusion (a.k.a. mutual- or cross-diffusion) where species movement results in opposite directions of the electrochemical potential of anions and cations, respectively. Here, the motion reflects the increase/decrease in the local anion or cation concentrations which is partially balanced by decrease/increase of the total charge concentration, as well as a change in solvent concentration. Significantly, these mechanisms are not introduced ‘by-hand’ to the model, they are results of the explicit inclusion of solvent and to a lesser extend the assumption of electrolyte incompressibility. Indeed, while Nernst-Planck dynamics arises from a random walk of isolated ions in an effective medium leading to Stokes-EinsteinSmoluchowski relations, the inclusion of solvent implies that species are not isolated. Thus, the movement of a cation, for example, involves pushing aside solvent molecules and possibly anions in his path and the movement of solvent to the void left behind him. At higher orders, the combined effect of all terms can be appreciated in one space dimension, by neglecting the boundary conditions and considering the case γ = Υ = 1. In this case, substituting (2b) into (5b), the flux takes the form of high order corrections for the drift Jp1 = p∇3 φ, and thus, ∇ · Jp1 gives rise to a fourth order derivative of the electric potential that addresses short range correlations 42–45 (we note that originally the high order operator was introduced phenomenologically through a displacement field in (2b)). In the next section, we show that, under appropriate conditions, these short-range interactions give rise to self-assembled nano-patterns near charged interfaces and/or in the bulk.

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1: Parameter space spanned by concentration (c) and ion size asymmetry (γ), showing the transitions from monotonic EDL as described by the PNP or Poisson-Boltzmann model (dark/blue shaded region), region of oscillatory EDL giving rise to under-screening (light/red shaded region), and region of self-assembled bulk. Parameters: Υ = 4, σ = 10, χ = 30.

Analysis: Non-monotonic screening length The broad family of generalized PNP models describes electrolyte solutions with a spatially uniform bulk. 20,30,46 Eqs. 4 gives rise to a richer picture of bulk and interfacial behavior, that is consistent with the picture arising in recent experimental and numerical studies. 21,23,33,35,47–49 Indeed, using spatial linearization methods, see SI for details, we map the qualitatively distinct bulk and interfacial behaviors, and the corresponding regions in the parameter space spanned by ion size asymmetry γ and normalized concentration c, see Figure 1. The results have been obtained in one space dimension (∇ → ∂x ) and with no-flux boundary conditions (inert electrodes) for ions and Dirichlet for the potential (constant applied voltage): J p (x = ±L/2) = 0, φ(x = ±L/2) = ±V /2, where L is the physical domain size and V is the applied voltage. Region A, describes the emergent self-assembly of bulk nano-structure, i.e., a spatially oscillatory profile throughout the whole domain, see top inset in Figure 1. Here, the region is obtained from the onset of the finite wavenumber instability, 38 see SI for details. This region is characterized by highly concentrated electrolyte with weak ion size asymmetries (typical size rations of roughly 1:2 or 1:3). Indeed, the first observation for self-assembly 10

ACS Paragon Plus Environment

Page 10 of 22

Page 11 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

was for protic IL with short alkyl groups, 33 although the phenomenon of self-assembly has a much broader context. 23 Region C, corresponds to a homogeneous bulk with a monotone electrical diffuse layer (EDL) structure, as shown in the bottom inset of Figure 1. This region is characterized by low enough concentrations so that steric effects are perturbative, and the model reduces to the PNP theory of dilute electrolytes (or Poisson-Boltzmann (PB) theory if steady states are of interest). The latter also demonstrates that sufficiently diluted IL, either by solvent 47 or possibly by formation of ion-pairs, 19,50 can behave as dilute electrolytes. The intermediate Region B, describes a homogeneous bulk with an oscillatory potential and charge density profiles near a charged interface (a.k.a, a spatially oscillatory EDL), as shown in the middle inset. The onset of this region is computed using spatial dynamics methods, 51,52 as detailed in the SI. Notably, these decaying oscillations have also been referred to as over-screening. 44,53 The three regions in Fig. 1, imply that, as an ionic liquid or concentrated electrolyte solution with relatively weak ion-size asymmetry (γ ≈ 1), is diluted, it exhibits a transition from nano-structured bulk and layered EDL at high ionic concentration (i.e., at region A) to layered EDL with homogeneous bulk (region B) and finally at low concentrations to a monotonic EDL structure. The insets A-C in Figure 2 present steady-state solutions of (4) with parameters corresponding to the ionic liquid, [C4 C1 Pyrr][NTf2 ], diluted with propylene carbonate (C4 H6 O3 ). Indeed, a transition from layered to monotonic EDL has been observed using high-energy x-ray reflectivity 47 and in situ AFM 48 within the range of parameters considered here. The molecular origin for the transition from monotonic to spatially oscillatory EDL, a.k.a. Kirkwood line, have been outlined using hypernetted chain 9 and mean spherical 54 approximations already in the early 1990s. However, as has been outlined by Smith et al. 15 and also demonstrated in Fig. 2, there are significant differences in comparison with empirical measurements. Moreover, the transition from region B to region A appears to be unexplored by available mean-field models. In what follows, we show

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

that this transition is associated with an unexpected decrease in the screening length with √ √ increasing concentration, 15,36 see region c¯ & 1.6 M in Fig. 2. To examine the transition from region C to B, we use the size symmetric γ = 1 case to derive analytic relations, which otherwise are obtained numerically, as shown in the SI. Above the monotonic to spatially oscillatory decay onset, the spatial decay is no longer being described by the PB approach and instead the screening length is better related to the envelope of the decaying oscillations. 9 In other words, the screening length is related to the spatial scale at which the bulk electroneutrality is violated 51,52 (tail of oscillation amplitudes) rather than to the electrolyte behavior in the vicinity of the solid surface. Indeed, in Fig. 2A, we observe that the envelope of the oscillatory tail (dashed curve) well describes the screening length, but does not describe the electrolyte behavior near the boundary, i.e., in the first two left spatial oscillations near x = 0. Following the re-definition of the screening length with the scale at which electroneutrality is departed, the formal computation involves linearization in space about a uniform bulk p ≡ p0 (c) and n ≡ n0 (c) corresponding to the normalized concentration c, p = p0 + δp eµx ,

n = n0 + δn eµx ,

φ = δφ eµx ,

where, |δp |, |δn |, |δφ |  1 are the respective eigenvector components at the onset. In this case, the screening length is the reciprocal of the real part of the spatial eigenvalues, λS ∼ Re(1/µ), that are associated with the eigenvalue problem, see SI for details. Particularly, in the sizesymmetric case (γ = Υ = 1) explicit expressions of these eigenvalues can be derived, where out of the several eigenvalues, the relevant ones are given by

µ2±

  q χc − 4 2 2 1 ± 1 − 16σc / (4 − χc) . = 2c2 σ

(6)

Figure 2 presents the computed screening length as a function of concentration for parameters corresponding to a specific ionic liquid, [C4 C1 Pyrr][NTf2 ], diluted with propylene carbonate. 12

ACS Paragon Plus Environment

Page 13 of 22

10 0.5

C

8

2

B

0

A

0

S

(nm)

0.5

Screening length,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

0

6

-2

-0.5 0

5

10

15

0

x

4

8

0 4 8

x

16

x

A

4

C 2

B 0 0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Figure 2: Screening-length as a function of concentration c¯ for ionic liquid, [C4 C1 Pyrr][NTf2 ], diluted with propylene carbonate (C4 H6 O3 ); the concentration is scaled according to dependence of Debye length. Experimental data from 15 (•) is superimposed with computed screening length (solid curve) for dimensional parameters that correspond to this ionic liquid, see SI for details. For comparison, we also plotted the monotonically decaying DebyeH¨ uckel approximation, λD (dotted red curve), the hypernetted chain correction 9 (×), and the curve 2/3¯ c3/2 (dashed curve) to allow comparison with the recently presented scaling 55 argument. The insets A-C present spatial profiles of the electric potential φ at different concentrations (solid curve), see (), and the exponential envelop (dashed curve) that is computed by the real part of the spatial eigenvalues (6), which is being used as a measure for the screening length. Dimensionless parameters: γ = 1, Υ = 1, χ = 28, σ = 35, V = 1 and L = 200. In accordance with experimental observations, at low concentrations where the EDL structure monotonically decreases, the screening length identifies with the Debye length. As elec√ trolyte concentration is increased above the Kirkwood point, cc = 4(χ − 4 σ)/ (χ2 − 16σ) ' 0.077, the envelope tail length increases and exceeds the Debye length. Respectively, in √ dimensional units the transition is at 0.48 M , which corresponds to the minimum in the IL screening length (Fig. 2). Accordingly, increase in the concentration leads to the phenomenon of under-screening. 15,36 As introduced by Lee et al., 55 in a regime of concentration beyond the Kirkwood point, we observe that the screening length roughly scales as c3/2 with respect to the Debye screening length. While the transition from regions C and B can be formulated using near equilibrium methods, it is not true for the transition from region B to A. Here, the significance of spa13

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

tiotemporal approach (2) in the context of screening length structure becomes fundamental as the temporal bulk instability is approached, i.e., region A. At the instability onset, the real part of the relevant eigenvalues Re(µ± ) → 0, and hence the screening length in a pure 1D system should extend to infinity, λS → ∞. This onset is in fact an analogue of the (temporal) finite wavenumber instability bifurcation point, as discussed in more detailed by Gavish et al. 37,41 However, in reality the system is even beyond the 1D representation, so that secondary time-dependent zigzag instabilities 56 of the bulk self-assembly become dominant in an appropriate parameter regime 37,41 . In this case, spatial oscillations of screening length are destroyed and the screening length shows a decreasing feature, cf. 40 Indeed, the second decay in screening length have been reported at high concentrations, see cf. 15,36 and Figure 2. The phenomenological origin of this behavior requires thus, a spatiotemporal framework as opposed to methods which focus on equilibrium behavior in regions B and C. 24,30,57,58 Consequently, as summarized in Fig. 2, the above results qualitatively capture the experimental observation of all three regions of the non-monotonic screening length dependence on concentration, as well as quantitatively agree with the PB theory. 15 We have also tested consistency of our results with parameters that correspond to aqueous NaCl, and obtained again a good qualitative comparison, as shown in Fig. 1 in the SI.

Conclusions In this work, we have introduced a spatiotemporal thermodynamically consistent continuum mean-field framework for electrolytes that ranges from solvent free to dilute concentrations with explicit treatment of the solvent medium and finite size (steric) effects of both ion and solvent components. These two components introduce, by consistency, an additional transport mechanism of inter-diffusion and bulk nano-structuring by self-assembly. We conjecture that these are the essential features that any mean-field for electrolytes should comprise to provide a comprehensive description at a wide range of concentrations that approach a

14

ACS Paragon Plus Environment

Page 14 of 22

Page 15 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

solvent-free case. However, in specific cases other approaches can be also efficient, for example, steric effects could be also introduced by using different approaches: a (repulsive) Yukawa potential, the hard sphere potential in Rosenfelds density functional theory of fluids, 30 and asymptotic PB-type models. 19,30,45,49,59,60 To emphasize that, we chose to focus on the screening length behavior at a full range of concentrations, which is beyond the available theories as it captures the effect of secondary transverse instabilities that explain the experimentally observed second decay at high concentrations. 15,36 Notably, the advantage of spatiotemporal framework is in cases where understanding of temporal and/or morphological bulk effects is required, such as self-assembly. A detailed analysis of secondary instabilities is beyond the scope of this letter and will be presented elsewhere. The advantage of the developed framework is not only consistency with different previous mean-field approaches 9,28,31,37,39,40,58,61 but it can be readily be extended to account for additional effects, e.g., ion-solvent interactions, ion-pairing, 19,22,62–64 or other electrostatic corrections. 65 These extensions should be advanced toward quantitative comparisons with experimental observations, beyond the qualitative trends depicted in see Fig. 2. Consequently, we expect that current spatiotemporal framework will stimulate experiments that combine temporal methods with structural analysis, 66 for example, transport properties such as multiple timescale relaxations 37 that have been already shown for pure ionic liquids due to inter-diffusion and saturated mobility 37 , have not been addressed here in detail, as systematic (for comparison purposes) experimental data is limited at large. 17,32

Acknowledgement This research was done in the framework of the Grand Technion Energy Program (GTEP) and of the BGU Energy Initiative Program, and supported by the Adelis Foundation for renewable energy research. N.G. acknowledges the support from the Technion VPR fund and from EU Marie–Curie CIG Grant 2018620.

15

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

References (1) Niedermeyer, H.; Hallett, J.; Villar-Garcia, I.; Hunt, P.; Welton, T. Mixtures of ionic liquids. Chemical Society Reviews 2012, 41, 7780–7802. (2) Perkin, S. Ionic liquids in confined geometries. Physical Chemistry Chemical Physics 2012, 14, 5052–5062. (3) Fedorov, M. V.; Kornyshev, A. A. Ionic Liquids at Electrified Interfaces. Chemical Reviews 2014, 114, 2978–3036. (4) Zhang, S.; Zhang, J.; Zhang, Y.; Deng, Y. Nanoconfined Ionic Liquids. Chemical Reviews 2016, 117, 6755–6833. (5) Armand, M.; Endres, F.; MacFarlane, D.; Ohno, H.; Scrosati, B. Ionic-liquid materials for the electrochemical challenges of the future. Nature Materials 2009, 8, 621–629. (6) Wishart, J. Energy applications of ionic liquids. Energy and Environmental Science 2009, 2, 956–961. (7) Salanne, M. Ionic Liquids for Supercapacitor Applications. Topics in Current Chemistry 2017, 3, 1–25. (8) Israelachvili, J. N.; Adams, G. E. Measurement of forces between two mica surfaces in aqueous electrolyte solutions in the range 0–100 nm. Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases 1978, 74, 975–1001. (9) Attard, P. Asymptotic analysis of primitive model electrolytes and the electrical double layer. Physical Review E 1993, 48, 3604. (10) Leote de Carvalho, R.; Evans, R. The decay of correlations in ionic fluids. Molecular Physics 1994, 83, 619–654.

16

ACS Paragon Plus Environment

Page 16 of 22

Page 17 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

(11) Biesheuvel, P.; Van Soestbergen, M. Counterion volume effects in mixed electrical double layers. Journal of Colloid and Interface Science 2007, 316, 490–499. (12) Mansoori, G.; Carnahan, N. F.; Starling, K.; Leland Jr, T. Equilibrium thermodynamic properties of the mixture of hard spheres. The Journal of Chemical Physics 1971, 54, 1523–1525. (13) Kirchner, B.; Malberg, F.; Firaha, D. S.; Holl´oczki, O. Ion pairing in ionic liquids. Journal of Physics: Condensed Matter 2015, 27, 463002. (14) Lee, A. A.; Kondrat, S.; Vella, D.; Goriely, A. Dynamics of ion transport in ionic liquids. Physical Review Letters 2015, 115, 106101. (15) Smith, A. M.; Lee, A. A.; Perkin, S. The electrostatic screening length in concentrated electrolytes increases with concentration. Journal of Physical Chemistry Letters 2016, 7, 2157–2163. (16) Yochelis, A.; Singh, M.; Visoly-Fisher, I. Coupling Bulk and Near-Electrode Interfacial Nanostructuring in Ionic Liquids. Chemistry of Materials 2015, 27, 4169–4179. (17) Makino, S.; Kitazumi, Y.; Nishi, N.; Kakiuchi, T. Charging current probing of the slow relaxation of the ionic liquid double layer at the Pt electrode. Electrochemistry Communications 2011, 13, 1365–1368. (18) Limmer, D. T. Interfacial ordering and accompanying divergent capacitance at ionic liquid-metal interfaces. Physical Review Letters 2015, 115, 256102. (19) Adar, R. M.; Markovich, T.; Andelman, D. Bjerrum pairs in ionic solutions: A PoissonBoltzmann approach. The Journal of Chemical Physics 2017, 146, 194904. (20) Bazant, M. Z.; Kilic, M. S.; Storey, B. D.; Ajdari, A. Towards an understanding of induced-charge electrokinetics at large applied voltages in concentrated solutions. Advances in Colloid and Interface Science 2009, 152, 48–88. 17

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(21) Igliˇc, A.; Drobne, D.; Kralj-Igliˇc, V. Nanostructures in Biological Systems: Theory and Applications; CRC Press, 2015. (22) Lee, A.; Vella, D.; Perkin, S.; Goriely, A. Are Roomerature Ionic Liquids Dilute Electrolytes? J. Phys. Chem. Lett. 2015, 6, 159–163. (23) Hayes, R.; Warr, G. G.; Atkin, R. Structure and nanostructure in ionic liquids. Chemical Reviews 2015, 115, 6357–6426. (24) Goodwin, Z. A.; Kornyshev, A. A. Underscreening, overscreening and double-layer capacitance. Electrochemistry Communications 2017, 82, 129–133. (25) Nernst, W. Die elektromotorische wirksamkeit der jonen. Zeitschrift f¨ ur Physikalische Chemie 1889, 4, 129–181. (26) Planck, M. Ueber die erregung von electricit¨at und w¨arme in electrolyten. Annalen der Physik 1890, 275, 161–186. (27) Bard, A. J.; Faulkner, Electrochemical Methods: Fundamentals and Applications, 2nd ed.; Wiley, New York, 2002. (28) Bikerman, J. Structure and capacity of electrical double layer. Philos. Mag. 1942, 33 . (29) Booth, F. The dielectric constant of water and the saturation effect. Journal of Chemical Physics 1951, 19, 391–394. (30) Varela, L. M.; Garcıa, M.; Mosquera, V. Exact mean-field theory of ionic solutions: non-Debye screening. Physics Reports 2003, 382, 1–111. (31) Horng, T.-L.; Lin, T.-C.; Liu, C.; Eisenberg, B. PNP equations with steric effects: a model of ion flow through channels. Journal of Physical Chemistry B 2012, 116, 11422–11441.

18

ACS Paragon Plus Environment

Page 18 of 22

Page 19 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

(32) Camci, M.; Aydogan, P.; Ulgut, B.; Kocabas, C.; Suzer, S. XPS enables visualization of electrode potential screening in an ionic liquid medium with temporal-and lateralresolution. Physical Chemistry Chemical Physics 2016, 18, 28434–28440. (33) Atkin, R.; Warr, G. The smallest amphiphiles:

Nanostructure in protic room-

temperature ionic liquids with short alkyl groups. Journal of Physical Chemistry B 2008, 112, 4164–4166. (34) Celso, F. L.; Yoshida, Y.; Castiglione, F.; Ferro, M.; Mele, A.; Jafta, C.; Triolo, A.; Russina, O. Direct experimental observation of mesoscopic fluorous domains in fluorinated room temperature ionic liquids. Physical Chemistry Chemical Physics 2017, 19, 13101–13110. (35) Pontoni, D.; Haddad, J.; Di Michiel, M.; Deutsch, M. Self-segregated nanostructure in room temperature ionic liquids. Soft Matter 2017, 13, 6947–6955. (36) Smith, A. M.; Lee, A. A.; Perkin, S. Switching the structural force in ionic liquid-solvent mixtures by varying composition. Physical Review Letters 2017, 118, 096002. (37) Gavish, N.; Yochelis, A. Theory of phase separation and polarization for pure ionic liquids. Journal of Physical Chemistry Letters 2016, 7, 1121–1126. (38) Bier, S.; Gavish, N.; Uecker, H.; Yochelis, A. From bulk self-assembly to electrical diffuse layer in a continuum approach for ionic liquids: The impact of anion and cation size asymmetry. Physical Review E 2017, 95, 060201(R). (39) Psaltis, S.; Farrell, T. W. Comparing charge transport predictions for a ternary electrolyte using the Maxwell–Stefan and Nernst–Planck equations. Journal of The Electrochemical Society 2011, 158, A33–A42. (40) Lazaridis, K.; Wickham, L.; Voulgarakis, N. Fluctuating hydrodynamics for ionic liquids. Physics Letters A 2017, 381, 1431–1438. 19

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(41) Gavish, N.; Versano, I.; Yochelis, A. Spatially localized self-assembly driven by electrically charged phase separation. SIAM Journal on Applied Dynamical Systems 2017, 16, 1946–1968. (42) Santangelo, C. D. Computing counterion densities at intermediate coupling. Physical Review E 2006, 73, 041512. (43) Liu, J.-L.; Eisenberg, B. Correlated ions in a calcium channel model: a Poisson–Fermi theory. Journal of Physical Chemistry B 2013, 117, 12051–12058. (44) Bazant, M. Z.; Storey, B. D.; Kornyshev, A. A. Double layer in ionic liquids: Overscreening versus crowding. Phys. Rev. Lett. 2011, 106, 46102. (45) Blossey, R.; Maggs, A. C.; Podgornik, R. Structural interactions in ionic liquids linked to higher-order Poisson-Boltzmann equations. Physical Review E 2017, 95, 060602(R). (46) Gavish, N.; Promislow, K. On the structure of Generalized Poisson–Boltzmann equations. European Journal of Applied Mathematics 2016, 27, 667–685. (47) Mezger, M.; Roth, R.; Schr¨oder, H.; Reichert, P.; Pontoni, D.; Reichert, H. Solid-liquid interfaces of ionic liquid solutionsInterfacial layering and bulk correlations. Journal of Chemical Physics 2015, 142, 164707. (48) Cui, T.; Lahiri, A.; Carstens, T.; Borisenko, N.; Pulletikurthi, G.; Kuhl, C.; Endres, F. Influence of Water on the Electrified Ionic Liquid/Solid Interface: A Direct Observation of the Transition from a Multilayered Structure to a Double-Layer Structure. Journal of Physical Chemistry C 2016, 120, 9341–9349. (49) Maggs, A.; Podgornik, R. General theory of asymmetric steric interactions in electrostatic double layers. Soft Matter 2016, 12, 1219–1229. (50) Gebbie, M.; Valtiner, M.; Banquy, X.; Henderson, W.; Israelachvili, J. Reply to Perkin et al.: Experimental observations demonstrate that ionic liquids form both bound 20

ACS Paragon Plus Environment

Page 20 of 22

Page 21 of 22 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

(Stern) and diffuse electric double layers. Proceedings of the National Academy of Sciences of the United States of America 2013, 110, E4122. (51) Yochelis, A. Transition from non-monotonic to monotonic electrical diffuse layers: Impact of confinement on ionic liquids. Physical Chemistry Chemical Physics 2014, 16, 2836–2841. (52) Yochelis, A. Spatial structure of electrical diffuse layers in highly concentrated electrolytes: A modified poisson-nernst-planck approach. Journal of Physical Chemistry C 2014, 118, 5716–5724. (53) Levin, Y. Electrostatic correlations: from plasma to biology. Reports on Progress in Physics 2002, 65, 1577. (54) Buzzeo, M.; Evans, R.; Compton, R. Non-haloaluminate room-temperature ionic liquids in electrochemistry - A Review. ChemPhysChem 2004, 5, 1106–1120. (55) Lee, A. A.; Perez-Martinez, C.; Smith, A. M.; Perkin, S. Scaling Analysis of the Screening Length in Concentrated Electrolytes. Phys. Rev. Lett. 2017, 119, 026002. (56) Cross, M.; Hohenberg, P. Pattern formation outside of equilibrium. Reviews of Modern Physics 1993, 65, 851–1112. (57) Attard, P. Electrolytes and the electric double layer. Advances in Chemical Physics 1996, 92, 1–160. (58) Storey, B.; Bazant, M. Effects of electrostatic correlations on electrokinetic phenomena. Physical Review E 2012, 86, 056303. (59) Ben-Yaakov, D.; Andelman, D.; Podgornik, R.; Harries, D. Ion-specific hydration effects: Extending the Poisson-Boltzmann theory. Current Opinion in Colloid & Interface Science 2011, 16, 542–550.

21

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(60) Giera, B.; Henson, N.; Kober, E. M.; Shell, M. S.; Squires, T. M. Electric doublelayer structure in primitive model electrolytes: comparing molecular dynamics with local-density approximations. Langmuir 2015, 31, 3553–3562. (61) Borukhov, I.; Andelman, D.; Orland, H. Steric effects in electrolytes: A modified Poisson-Boltzmann equation. Physical Review Letters 1997, 79, 435–438. (62) Marcus, Y.; Hefter, G. Ion pairing. Chemical Reviews 2006, 106, 4585–4621. (63) Merlet, C.; Madden, P. A.; Salanne, M. Internal mobilities and diffusion in an ionic liquid mixture. Physical Chemistry Chemical Physics 2010, 12, 14109–14114. (64) Hollczki, O.; Malberg, F.; Welton, T.; Kirchner, B. On the origin of ionicity in ionic liquids. Ion pairing versus charge transfer. Physical Chemistry Chemical Physics 2014, 16, 16880–16890. (65) Schlumpberger, S.; Bazant, M. Z. Simple Theory of Ionic Activity in Concentrated Electrolytes. arXiv:1709.03106 [physics.chem-ph] 2017, (66) Reichert, P.; Kjær, K. S.; van Driel, T. B.; Mars, J.; Ochsmann, J. W.; Pontoni, D.; Deutsch, M.; Nielsen, M. M.; Mezger, M. Molecular Scale Structure and Dynamics at an Ionic Liquid/Electrode Interface. Faraday Discussions 2017,

22

ACS Paragon Plus Environment

Page 22 of 22